AMZN FY25 result--top 10 holding--AMZN are going all in. May be a bumpy ride.

February 09, 2026

AMZN FY25 RESULT

Very solid result from AMZN with the numbers modestly above my estimates, with revenues +12% and NPAT +31%. There was a large MTM amount for the Anthropic investment of about $13.3B pre tax. If we remove that, NPAT growth becomes 15%. Overall, good results. There were also various other negative one-offs, including a $1.1B tax charge in Italy, $730m severance charges and $610m impairments in NA stores. Most of these were in the retail section.

Retail

NA sales were 10% higher and operating income 18% higher, International 13% and 26% respectively. These numbers included the one-off charges mentioned above. Advertising continues to drive results as AMZN has a structural advantage in having direct intent of customers when placing ads. Incremental ad revenues were $12.4B, compared to combined operating income growth of $5.5B.

AMZN continue to promote product range, speed of delivery and sharp pricing as features of the business, together with prime membership. Third-party sales continue to inch higher (64% 25, 63% 24) in the mix, which is positive for profitability.

Rufus, the agentic shopping tool, can now move to third-party sites, and AMZN continues to promote its efficiency against AI crawlers. As for AI crawlers, AMZN continue to be critical of them, but appears to be willing to do a deal, allowing AI crawlers onto the AMZN site, if an appropriate profit share is reached.

Tariffs are negative for AMZN, but in a relative sense, AMZN is not as exposed to de minimis rules as Shein and Temu are, so a relative gain.

AMZN has made significant gains in “Everyday Essentials”, groceries.

Disappointingly, not much was made of efficiency gains from AI and robotics in retail. Although there is some sensitivity here, and there have been announcements of large staff cuts, margin gains from efficiency remain a core thesis. That said, efficiency measures continue to improve for the whole group.

AWS

Revenues were 22% higher and operating income 15% higher, so a hit on margins which became apparent throughout the year as investment costs and wage costs both were higher.

Management consistently stated that demand exceeds supply and that AWS is monetising capacity as quickly as it can be deployed. Non-AI workloads from premises to cloud continue to grow strongly. AWS has invested in its proprietary chips with the aim to reduce costs to serve. A large slew of products designed to allow customers to access LLMs, fine-tune and build their own models, integrate them, and build and use agents are being developed or deployed. AMZN commented that outside clients are larger than internal clients. AMZN is also a believer in the diffusion of AI, that the current use cases and client lists will grow enormously over time, and now is the time to build the asset base and is confident of strong ROIC on this investment.

Other

The higher margin segments, AWS, ads and subscriptions continue to grow faster than the lower margin segments. That mix change is a positive for profitable growth.

A huge capex number $200B was touted, the spend was outlined as targeted on AI, chips, satellites, quick commerce and everyday essentials. Most targeted at AI.

Project Leo, the satellite business, is due to launch this year. Already 180 satellites are in orbit, another 20 are due to be launched in 2026 and 30 in 2027. A $1b increase in costs was forecast.

First quarter guidance was positive at the top line, being 15% growth, but at operating margins of about 10.8% as costs of expansion on various fronts hit the income statement.

SUMMARY AND VALUATION

AMZN is one of the most difficult large stocks. The thesis is quite simple: the domination in e-commerce will be enhanced by AMZN’s lead in AI and robotics, and margins will ultimately move higher in retail. AWS's business will ultimately develop into a lucrative oligopoly with the other hyperscalers. The inherent cash flow and margins are lower than those of some of the other players and are exposed to more cyclical retail. Together with such large investments, capital employed up 30% cagr over the last five years, bringing about a large chance of variability in reported results. Continual growth weighs on short-term profitability, and the market will lose patience.

Compared to the other large US tech companies, AMZN has always been willing to aggressively grow capex and push returns into the future. That long term wealth creation, if executed correctly, brings reasonable returns over the long term, but with a higher degree of volatility. Thanking aloud, that brings maybe more movements of the weight of the stock in the portfolio.

My broad assumptions on 5Y growth are NA retail 8% cagr and an exit margin and PE of 8% and 21x, International retail, 10% with an exit margin of 7% and a 22X PE. Note that international is a mix of highly profitable mature businesses and loss-making start-ups. As growth slows, profits will occur, so it’s a growth/profit trade-off. AWS to grow 15% compound, and a 23X exit PE. At $223 gives a 8% return. Returns above 10% are generated below $200, and there is a chance that if conditions tighten, investors will get a shot at a lower SP, due to higher risks.